Lowering the cut off value of an automated chlamydia enzyme immunoassay and confirmation by PCR and direct immunofluorescent antibody test.

Open Access
Publisher Website
Google Scholar
Abstract
To increase the sensitivity of an automated chlamydia enzyme immunoassay by significantly lowering its cut off value, and to maintain specificity by confirmation with polymerase chain reaction (PCR) and direct immunofluorescent antibody test (DFA). Over five months, the cut off value of the enzyme immunoassay used to screen urogenital samples for chlamydia antigen was reduced from 80 to 10. Samples with a test value of 10 or above were further tested with a commercial PCR assay. All samples during the first three months and discrepant samples during the last two months of the study were also tested with the DFA. 3250 urogenital swabs (1246 urethral, 1335 endocervical, 669 pooled urethral/endocervical) from 1246 males and 2004 females were processed. Using the manufacturer's recommended cut off of 80, the enzyme immunoassay identified chlamydia antigen in 134 samples (4.1%). Using the lower cut off value of 10 and either PCR or DFA as the confirmatory test, Chlamydia trachomatis was identified in 178 samples (5.5%). Thus, 45 additional positive samples were identified and the confirmed detection rate was increased by 33.8% (45/133). Excluding equivocal PCR results, the concordance between DFA and PCR was 91.8%. This strategy increased the detection rate by 2.1% in men and 0.9% in women (significant only in men). In female patients, pooled urethral/endocervical swabs as a specimen gave a significantly higher yield than endocervical swabs regardless of whether the lower cut off strategy was used. This strategy of significantly lowering the cut off test value with confirmation on the same specimen by either PCR or DFA is feasible and cost effective. The use of pooled urethral/ endocervical specimens in females should be considered routinely as detection rate was significantly improved.

This publication has 12 references indexed in Scilit: